Figure 2
Classification of calls to the valence and context of production based on t-SNE. t-SNE embedding of (a) valence (embedding perplexity = 50) and (b) context (embedding perplexity = 20) classifying neural network’s last fully connected layer activations for each spectrogram (t-SNE plots visualize the probability that two points are neighbors in an original multivariate space). Triangles indicate negative valence vocalizations, while circles indicate positive ones (see Supplementary Text for more information on the settings used for this figure).
